INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 2020 ISĀ©SN 2277-8616 335 IJSTR©2020 www.ijstr.org An Integrated Biostatistical Approach To Reveal The Health Status Among Elderly People At Receiving Home Care Wan Muhamad Amir W Ahmad, Muhammad Azeem Yaqoob, Rabiatul Adawiyah Abdul Rohim, Farah Muna Mohamad Ghazali, Nor Azlida Aleng Abstract: This paper examines the factors influencing the health status among the elderly at Rumah Seri Kenangan (RSK), Pengkalan Chepa, Kelantan and RSK Bedong, Kedah. Correlation Analysis (CA), Decision Tree Analysis (DTA), Multilayer Perceptron (MLP), and Principal Component Analysis (PCA) were used to determine the factor that might be associated with the health among the elderlies in both RSK. Through these methodologies, the health status factor will be assessed and validate simultaneously. Results from these analyses will be used as a benchmark for the decision making especially among the decision-maker to improve the level of quality which given to the elderly. The utmost finding from this study, it provides very useful information to the health caregiver for future management action plan and to improve the existing management system of an elderly. Index Terms: Multilayer perceptron, principle components analysis, correlation analysis, and decision tree analysis. —————————— ◆ —————————— 1. INTRODUCTION HEALTH status includes physical, social, and mental health. Assessment of disease, such as signs, symptoms, and physiological stress measures, and determine the illness, like functional status, are embedded in the concept of health status [12]. Assessment of the health status has been suggested as a pivotal health determinant, specifically in primary care centre, with preference being given to health promotion and prevention [6]. Even though aging is an extremely individual process which effects the health status of elder individual, there is copious evidence that their health status is correlated with a combination of risk factors of recession in functional status, like psychological stress, comorbidities, cognitive impairment, smoking, less physical activity, high body mass index (BMI), and less social contact [14].Psychological stress is a pathological process in elderly people, not a physiological reaction to growing elder. The mostly people confront with ageing, and many feel glad and satisfied. However, there is a not agreement among health experts and the society in broad spectrum to accept reduced functioning and high incidence of symptoms in elder people [1]. It is publicly accepted that elderly people have high burden of depression, but some studies shown younger have higher level of psychological stress [2]. Identification of elderly people with high and low risk for prospective dementia has appeared as an essential clinical and public health issue [7]. To address these concerns, we assessed health status which includes the psychological stress, neuropsychological disorders and other factors in elderly individuals in Kelantan, Malaysia. Many studies had been conducted especially on improving the existing management of the elderly. Most of the study is emphasizing on the factor related to health care. There are many statistical analysis tools that had been used to determine the factor as such Multi-layer perceptron (MLP), Principal Component Analysis (PCA) and Correlation Analysis (CA) and Decision Tree Analysis (DTA) and many more. The artificial neural network paradigm has systematically demonstrated its efficacy as a reliable nonlinear classification technique [10]. Multi-layer perceptron (MLP), is a class of feedforward artificial neural network (ANN) used for data classification, requires the class labels needed for each sample to be compared to the actual output produced by MLP [9]. The term MLP is used vaguely, sometimes loosely to refer to feedforward ANNs, sometimes strictly to refer to networks consisting of multiple layers of perceptrons (with threshold activation) [5]. MLP consists of at least three layers of nodes: an input layer, a hidden layer, and output layer. Except for input nodes, each node is a neuron that uses a non-linear activation function. MLP uses a supervised learning technique called backpropagation for training [11] [13]. Multiple layers and non-linear activation distinguish MLPs from linear sensing. It can distinguish data that cannot be separated in a linear way [3]. Principal component analysis (PCA) is a mathematical procedure that converts several numbers of (possibly) correlated variables into a (smaller) number of uncorrelated variables called principal components. PCA calculates a small number of orthogonal directions that contain most of their variability. Proper solutions for PCA have long been used [8]. PCA on correlation is much more informative and reveals some structure in the data and relationships between variables. Correlation analysis is a statistical method used to study the strength of the relationship between two, numerical and continuous variables such as height and weight. This particular type of analysis is useful when wants to determine if there is a possible relationship between variables [4]. 2 MATERIAL AND METHODS The source of population comprises of an elderly which age more than 60 years old and living in Rumah Seri Kenangan (RSK) in Pengkalan Chepa, Kelantan and RSK Bedong, Kedah. RSK is government funded public sheltered home for elderly suffered from lack of financial and family support. ———————————————— • Wan Muhamad Amir W Ahmad, School of Dental Sciences, Health Campus, Universiti Sains Malaysia, Malaysia. E-mail: wmamir@usm.my • Muhammad Azeem Yaqoob, School of Dental Sciences, Health Campus, Universiti Sains Malaysia, Malaysia. E-mail: dr.axeem.sr@gmail.com • Rabiatul Adawiyah Abdul Rohim, School of Dental Sciences, Health Campus, Universiti Sains Malaysia, Malaysia. E-mail: adawiyah5350@yahoo.com • Farah Muna Mohamad Ghazali, School of Dental Sciences, Health Campus, Universiti Sains Malaysia, Malaysia. E-mail: muna_ghazali@yahoo.com • Nor Azlida Aleng, Faculty of Ocean Engineering Technology and Informatics, Universiti Malaysia Terengganu, Malaysia. E-mail: azlida_aleng@umt.edu.my